Even genuine replacement Apple displays can mess with iPhones


Following the news that the latest iOS update can break phones with non-official replacement screens, repairers are encountering a different, more subtle problem: If you put a genuine Apple replacement display into an iPhone 8, 8 Plus or X, it'll no longer be able to adjust its brightness automatically. If Apple or one of its authorized partners were to put the same display in the same phone, though? The aftermarket repair community has verified the behavior in phones from the US all the way to Australia. It's confirmed to be an issue with phones running iOS 11.1, 11.2 and 11.3, which led sources to suggest it's been a problem since the launch of the latest batch of iPhones last fall. I was able to confirm that even swapping the displays of two brand-new iPhones causes the ambient light sensor to stop working, despite it not being altered or touched in any way.

Under the bed motion sensor lights are genius


Just to let you know, if you buy something featured here, Mashable might earn an affiliate commission. Unless your bed is right next to your light switch, you'll understand the struggle of trying to get up in the dark and tripping over everything in the room until you can get the light on. This makes getting up for a midnight snack or to use the bathroom in the middle night a total pain. Thankfully, under the bed motion sensor lights are a thing and a surprisingly simple solution to this issue. These LED strip lights are 3 meters (almost 10 feet) long and have adhesive on the back.

Samsung's ISOCELL Plus camera sensor upgrades low light performance


While Samsung may be playing catch up in some fields, it continues to charge ahead with its smartphone camera tech. Today it's unveiled its new ISOCELL Plus technology, which means sharper and more accurate photos even in challenging light environments. ISOCELL Plus, unsurprisingly, follows on from its first iteration of ISOCELL back in 2013, followed by last year's focus-boosting imaging sensor that could slot neatly into ultra slim phones, and then the 16-megapixel Slim 3P9 launched earlier this year, which also boasted a plug-and-play solution for mobile devices. ISOCELL technology works by creating a physical barrier between pixels, which reduces color crossover and allows each pixel to absorb more light than the typical backside-illuminated image sensors. However, as this barrier is made up of a metal grid, some light can be inadvertently absorbed or reflected.

Evidence with Uncertain Likelihoods

arXiv.org Artificial Intelligence

An agent often has a number of hypotheses, and must choose among them based on observations, or outcomes of experiments. Each of these observations can be viewed as providing evidence for or against various hypotheses. All the attempts to formalize this intuition up to now have assumed that associated with each hypothesis h there is a likelihood function {\mu}h, which is a probability measure that intuitively describes how likely each observation is, conditional on h being the correct hypothesis. We consider an extension of this framework where there is uncertainty as to which of a number of likelihood functions is appropriate, and discuss how one formal approach to defining evidence, which views evidence as a function from priors to posteriors, can be generalized to accommodate this uncertainty.

Inverse Filtering for Hidden Markov Models

Neural Information Processing Systems

This paper considers a number of related inverse filtering problems for hidden Markov models (HMMs). In particular, given a sequence of state posteriors and the system dynamics; i) estimate the corresponding sequence of observations, ii) estimate the observation likelihoods, and iii) jointly estimate the observation likelihoods and the observation sequence. We show how to avoid a computationally expensivemixed integer linear program (MILP) by exploiting the algebraic structure of the HMM filter using simple linear algebra operations, and provide conditions for when the quantities can be uniquely reconstructed. We also propose a solution to the more general case where the posteriors are noisily observed. Finally, the proposed inverse filtering algorithms are evaluated on real-world polysomnographic dataused for automatic sleep segmentation.